Whenever we add effects to sounds within a mix, those effects tend to fall into one of two categories: insert effects to individual sounds or auxiliary effects that can be shared by a number of sounds at once. Broadly speaking, EQ and compression are best suited to the former, while reverbs and delays, which we'll usually want to access for a number of different sounds, work best as auxiliaries.

This tutorial, however, focuses on parallel processing techniques, which form a grey area between these approaches. It's worth remembering that when we set up a reverb effect via an auxiliary, the whole of our dry sound continues to play back via its own channel, with a second channel of reverb mixed in; these twin channels combine in the output to produce a mix between the two. While we don't tend to refer to reverb effects set up this way as a parallel process, this is exactly what we've created: two channels running side by side to produce a sonic result we like. Parallel processing can be used to create all kinds of sonic treatments, however, and through this tutorial, we'll discover that swapping the reverb effect in the example given above for other effects we'd normally use as inserts—EQ, compression and distortion—can add powerful mix weight to our tracks.

Programmed beats: Let's start by looking at a parallel compression technique on drums, to bring extra power to some programmed hits. Here, I've programmed a simple pattern spread over three tracks. The first is a kick drum, the second a snare and the third a hi-hat pattern. Here are the beats with no additional processing.

Parallel compression setup: The pattern is very simple and, of course, at the moment, these drums aren't being balanced against any other mix elements. But they still sound quite thin and lacking in weight. We could add individual track processing to each sound to give them more substance, but instead we're going to send all three parts to Auxiliary 1, where we're setting up the FabFilter Pro-C Compressor. I want to add power to the beats, so I'm dropping the threshold to a low level to catch the entire input signal, before selecting the classic compression algorithm and setting a high ratio of 7.8:1 to squeeze the beats hard. Setting short attack and release times sounds best, giving the compressed beats a pumped feel. In this clip, you can hear just the parallel compression treatment soloed, without hearing the dry drum sounds at the same time.

Parallel-processed beats: As you can hear, the resulting sound is pretty smashed, and if this were our only drum source, the compression level would be overpowering in most mix contexts. However, when we mix the parallel treatment back in with the original drums, we gain the unprocessed impact of the dry drum sources, with their sharp and natural attack portions (called transients), as well as the super-compressed drum bus, leaving us with a much more powerful drum sound. Compare this audio file with the first one to hear the difference.

Of course, we have level control over this parallel treatment, too. If the compressed drum treatment is dominating the dry drum sources, we can simply drop the fader level of the auxiliary bus to more subtly blend these reinforced drums under the original programmed beats. That, of course, will depend on the requirements of the mix, which are impossible to judge until our track has been fleshed out to a fuller arrangement.

Track plus original beats: Let's do that now. In this clip, you can hear that the track has gained other elements. There's a driving bass sequence as well as another sub-bass, which is using the modwheel to create growling tone ramps through each bar. There's also a heavily delayed lead sound and a gentle pad. In this clip, you can hear these elements playing against the original beats without their parallel treatment. As you can hear, the drums are now drowned—the energy and extended frequency range of the new sounds completely dominates the original programmed drums.

Track plus full-level parallel: Let's mix in the parallel beats at their full original volume. The power of the compressed treatment should instantly give the beats considerably more drive and mix weight.

Track plus balanced parallel: That's exactly what we get—rather than being relegated to the bottom of the mix where they have no impact, the drums now sound richer, fuller and punchier. They're probably too loud now, in fact. So, to blend this parallel treatment to a more appropriate level, all we need to do is drop the fader level of the auxiliary bus where the parallel compressor is inserted and put it at a level where it will be less domineering. We settle for a fader position of -4.2 dB.

Of course, the level of the parallel channel may need to change as the track undergoes further development; if the arrangement suddenly expands to include a range of more powerful sounds, we may need to increase its volume, whereas if it drops back to a sparser arrangement, less level may be required. It's common for mix engineers to use extensive automation on parallel channels to ensure that they're at the right level. In whichever DAW you work with, don't be afraid to create volume automation lines to keep parallel channel volumes under control.

Less hats: There are further considerations for our parallel treatment, too, relating both to volume and tone. The first is that, at the moment, all of our drum sounds are feeding into our parallel channel at the same volume, as the send levels from the kick, snare and hat parts are all identical. As a result, the compressor is responding to a perfect copy of the dry levels of those sounds, whereas a slight variation might produce better results. For instance, the hats level, after the compressor, seems too loud compared to the volume of the kick and snare, and as there is a small reverb treatment coming from the hat part, its insistence in the mix is compounded by the parallel treatment. To lessen this, we can back down the send level of the hats to the parallel compressor to achieve a better balance. In this clip, you can hear the parallel treatment at its full volume under the dry drums, followed by the same treatment with less send level from the hats part. Finally, you can hear this new hats balance in the context of the whole track.

Parallel compression with EQ: That definitely works better, but what's also clear is that the reinforcement of the kick drum may also be a little loud. We could adjust this by again dropping the send level from the kick to the parallel channel, but we could also address it by adding a second effect after the compressor here. The dry kick is providing the main thump, so by inserting an EQ after the compressor we can tailor the additional kick weight to balance it against the rest of the track. I'm using DMG's Equilibrium EQ with a low shelf roll-off below 64 Hz to scoop the lowest frequencies by 6.5 dB. Then a bell curve is introduced at 133 Hz to knock the fundamental frequency of the kick down by 2.7 dB. I've also taken the opportunity to boost the body of the snare a little, at 489 Hz by 1.5 dB. Here, you can hear the original parallel treatment on the drums, without EQ, before hearing the same drum part with this EQ involved. Finally, you can hear how this treatment sounds in the context of the whole track.

What's clear is that just like regular inserts on individual sounds, we can construct chains of effects on our parallel channels to compose the best possible treatments for our tracks.

Track start: And the example above shows us that EQ can be used as a hugely useful parallel effect. In the next example, we've got a track at 100 BPM featuring a main beat loop, a bassline and a filtered pad. This last element has an insert compressor sidechained from a kick drum source. This trigger is silent, so we don't hear the kick within the track, but it ducks the compressor on each beat of every bar, giving a volume-sucking quality to the pad. The final element of the mix is a second beat loop, which features a ride cymbal as its most prominent instrument. This part is being treated with a ramped filter plug-in so that over every two bars the pattern goes from bright to dull, with a generous resonance amount ensuring that this loop sounds quite fizzy as the filter closes.

The pad sound is quite dull and muted at the moment, as it's being filtered. However, the core sound works well, so we don't want to open the filter to change the frequency balance of this. However, we could try enhancing the pad with a second tonal treatment to breathe some air and life into it.

Pad with parallel EQ: To do this, we're going to set up a parallel EQ treatment on Auxiliary 3. There, we're using FabFilter's ProQ 2 and being pretty brutal, with a low shelf to roll off frequencies below 1.16 kHz by 18.5 dB and a high shelf to boost frequencies above 2.8 kHz by 11 dB. This produces a super-bright treatment that would be far too thin and harsh if it were an EQ curve used over the original pad. But as a parallel treatment used in addition to the original, much more air and life is added to the pad. You can hear the pad soloed here, with the original playing first, then the blend of the original and the parallel EQ treatment playing afterwards.

Track with parallel EQ: Let's mix the parallel treatment in with the track. The parallel channel brings the pad up into the mix, adding a kind of celestial quality to this mix element, particularly when you compare it to the original-mix clip above.

Autopanning parallel EQ: However, this new high frequency content is now fighting the filter movement on our filtering drum loop. One way to stop this conflict and to ally the parallel EQ to this part would be to copy the auto-pan setting from the drum loop to the new parallel EQ channel. As well as filtering down over two bars, the drum loop is also moving from one side of the mix to the other, using SoundToys' PanMan, so copying this insert effect to the parallel EQ channel should help glue the additional air in the pad to this drum loop.

Of course, if that stereo movement is too stark, we could smudge it by sending the parallel EQ treatment to the delay on Auxiliary 2, which adds a further step of mix complexity. So now we know that not only can different effects be used as parallel treatments, and that those effects can be further enhanced by longer effects chains, we also know that we can route our parallel auxiliary treatments into other ordinary auxiliary treatments such as reverb and delay.

In our final example, we're going to look at using distortion as a parallel treatment. For a long time, distortion was seen as a guitar-only effect, but, due to the grit it can provide, it's become an essential addition to signal chains for electronic producers of many genres. As we'll soon see, even at low levels, various forms of distortion can be fantastic effects to add as parallel treatments to lead vocals. Distortion expands the harmonic range of input signals, so if your vocal is soft and floaty, for example, reinforcing lower frequency content with parallel distortion can be hugely effective.

Vocal track start: Here's the track we'll work with, which is deliberately sparse, as the lack of density in the backing track will showcase how even a small amount of parallel distortion can make a big difference. The track features a set of percussive sounds to provide a rhythmic foundation, while a sub-bass underpins a pad part. The lead vocal has been copied so it plays twice. The copy is a 16th note late (to provide a kind of echo effect), and its volume is significantly reduced.

Parallel distortion: The vocal is extremely soft—it's been close mic'd and recorded softly to match the emotion and flavour of the track. But while its airy, light quality adds a vulnerability to the performance, it lacks the weight and brooding feel of the pad and percussion hits. To bring this quality to the vocal, we set up SoundToys' Decapitator on Auxiliary Bus 3 and add plenty of gain, and use the tone control to slightly favour a dark rather than light character. Dropping the parallel distortion channel fader to -5 dB lets you hear this effect easily, but it's far too overwhelming. During the first vocal pass, the track plays without parallel distortion, and it's then introduced for the second pass.

Less parallel distortion: Of course, the parallel channel completely overwhelms the mix. But if you can look past how loud it is, you can hear that the tone of this treatment is valuable—it creates bite and drama that the original vocal doesn't have. So it's now just a question of balancing in the right amount of it. And it's a huge surprise to discover just how quiet this treatment can be to still make a positive difference to the mix. Here's the same track with the parallel distortion fader dropped to -17.6 dB. To give an idea of how subtly effective the distortion channel is, I've drawn automation across the first vocal pass so that the first and third phrases feature the distortion, while the second and fourth don't. Then, through the second half, the treatment is present the whole way through.

This shows us that parallel treatments can play a vital role even when set deep in a mix. As always, it all depends on context, and your ears informing the choices you make. Now that you've seen how parallel treatments are set up, you can explore your own approaches to techniques for compression, EQ and distortion and any number of other creative applications.